We brought an entire month of by-elections to a close with the two Perth City by-elections in Perth & Kinross. These are also the last by-elections of 2020 – no others due in Scotland, Northern Ireland typically fills vacancies by co-option anyway, and votes remain suspended in England and Wales until May. That means the 8 Scottish by-elections in October and November have been the only elections held in the UK since early March!
Even before the votes were cast, we knew this double-dunter was going to be a Tale of Two Perth Cities. In the North, the SNP had come close to half the vote in 2017, and I didn’t think they’d have any competition this time around.
The South, however, was set to be a close three way marginal between the SNP (who caused the vacancy), the Conservatives (who won a 2017 by-election in the ward), and Lib Dems (who won the most votes in the full 2017 election.) Though I had that one figured as a tossup between all three parties, I did think the Lib Dems had more paths to victory, with transfers clearly favouring them if they placed either first or second.
First Preferences - South
As it’s the more exciting of the pair, let’s look first at Perth City South. Not to toot my own horn too much, but my assessment of the possibilities there was spot on. First preferences in full were (note changes are vs 2017 by-election / 2017 full election):
SNP - 1898 (32.9%, +0.8 / +7.1)
Liberal Democrat - 1823 (31.6%, +2.8 / -3.1)
Conservative - 1698 (29.4%, -1.8 / +4.2)
Labour - 204 (3.5%, -2.1 / -2.8)
Green - 135 (2.3%, +0.5 / -0.7)
UKIP - 18 (0.3%, +0.3 / +0.3)
Note that in the 2017 by-election an Independent won 0.5% of the vote, and in the full 2017 election two Independents won 5.0% between them.
Looking at the changes versus the by-election as being most salient, none of these are particularly big. Just under 3% gain for the Lib Dems is the largest gain, and just over -2% loss for Labour the sharpest drop. Given how fine the margins were at that by-election though, this was enough for the Lib Dems to push into second ahead of the Conservatives. That put them just 75 votes behind the SNP, though all three parties remained very close to one another, with just 3.5% between the SNP in first and Conservatives in third.
In terms of comparing with the full election in 2017, much larger gains for the SNP and Conservatives are evident. Those will be in part down to the absence of Independents, and in part due to a slight redistribution of what was evidently a strong personal vote for the local Lib Dem councillor elected then. Labour were down on both measures here, continuing a slow decline, whilst the Greens were up versus the by-election but down versus the full.
In my preview, I’d said that if the Lib Dems came second, they’d almost certainly win given both the SNP and Conservatives prefer them to one another. When the transfers fell, that’s exactly what happened.
Transfers - South
Looking now at the transfers, and taking us up to the final stage 6 (changes vs final head-to-head stage in 2017 by-election / 2017 full election):
Lib Dem - 2749 (47.6%, +47.6 / -8.3)
SNP - 2110 (36.5%, -3.6 / +8.6)
Non-Transfer - 917 (15.9%, -1.1 / -0.3)
What was narrowly trailing became a comprehensive Lib Dem victory once the Conservatives were eliminated. Again, however, note that compared to the full election in 2017 there was a weakening of the Lib Dem vote and strengthening of the SNP’s. That doesn’t detract from Liz Barrett’s victory here, it’s just a reminder this result is more in line with recent polling than the flipped councillor would suggest.
Indeed, this is a really good example of why here at BBS I generally avoid the terminology of “Gain/Loss/Hold”. It’s certainly true that in practical terms of seats on the council, this is a Lib Dem gain from the SNP. But in terms of previous election results, it’s a gain from the Conservatives (versus the 2017 by-election), or a hold (versus the 2017 full election). We’re very used to FPTP meaning that “who held the seat” versus “who had the most votes” are equivalent measures, and the fact they aren’t under STV can be confusing.
In-Depth Data - South
As this was a machine count, there was a whole bunch of juicy additional data available. I’ve been quite bad at coming back and adding these recently because the data isn’t always released particularly quickly. I’m going to try and get in the habit of doing results posts on Monday or Tuesday nights to give time for data release, rather than the Friday rush I’d settled into.
Anyway, the first of our fun bits of extra data is looking at where each party’s second preferences go. Remember we can’t learn this just from transfers, because as soon as you eliminate someone, they can’t get any second preferences they are down for when anyone else drops out later.
Some of the usual stories here, with strong mutual SNP-Green preferencing, whilst Conservatives and Labour look a bit more to the Lib Dems. That massive Con to Lib Dem preference is in large part what made the latter’s win so comprehensive here.
Interestingly, there’s a stronger favourability in the other direction here than there has been in other recent by-elections. However, it wouldn’t have been quite enough to close the gap had the Lib Dems dropped out before the Conservatives, as that would have been 42.4% SNP vs 41.5% Conservative.
Then we can look at how support was distributed throughout the ward.
The main change here versus the 2017 by-election is that the Lib Dems come ahead in the PLD district covering, amongst other areas, Cherrybank. Last time it had been the Conservatives in the lead there. We can tell that a lot more voters live in and/or turned out in these orange bits of the map, because the Lib Dem lead over the SNP in them isn’t always enormous, but they were very close in total votes across the ward. By contrast the SNP tended to have larger leads in the districts they topped, in particular in Letham where they won almost 60% of the vote, which will be why they led overall.
With the Conservatives no longer leading in any district, their best showing was now in the combined PLA & PLB district in the south of Moncreiffe. And though heavily depleted beyond even what it was then, Labour’s strongest result was in the chunk of Letham that’s in this ward.
Phew! That’s a lot of information for one ward… so let’s do it all again for North!
First Preferences - North
Again, the expectation for this one was entirely met with an SNP win. First preferences in full were:
SNP - 1406 (61.0%, +12.8)
Conservative - 528 (22.9%, -2.5)
Labour - 220 (9.5%, -6.2)
Liberal Democrat - 91 (3.9%, +0.5)
Green - 60 (2.6%, +2.6)
Note that in 2017 three Independents won 6.6% of the vote between them.
A stonking victory on first preferences alone here, becoming the third time any party has achieved that since I launched BBS in January 2018. All three of those were this year, most recently last week in Clackmannanshire for the Conservatives!
Those big SNP gains came largely at the expense of Labour, which although it had a much better day here than in South still meant losing over a third of their vote share. That doesn’t bode well for the party’s chances of regaining a councillor here in 2022. The Conservatives too had a small drop in support, whilst a small uptick for the Lib Dems and the first appearance for the Greens accounted for the rest of the changes.
In-Depth Data - North
See above for what this section all means – no use repeating myself!
The main story here is just how few transfers there would have been had they been necessary. It’s not entirely unheard of for Conservative voters to do a single preference vote, but for that to be true of the SNP and nearly for Labour as well is quite something. What’s probably most notable here is that given the makeup of the ward, the Lib Dems weren’t as friendly to Conservatives here, with a relatively even split between all parties but particularly favouring Labour.
Again, next section looks at how support was spread around the ward.
Given the SNP had the lead everywhere in 2017, it won’t be a surprise to see that remains the case in this blowout. They even come within touching distance of three-quarters of all in-person votes cast in the Letham section! Everyone else has their best result in the Hillyland area, which is therefore the only district not to have a majority vote for the SNP.
Anyway, that’s us done with by-elections for the year! Though the pandemic has done and continues to do much worse, it is unfortunate to be missing out on a handful more by-elections that might have squeezed into December otherwise. As in previous years I’ll do a post reflecting on them as a whole at some point in December.
If you find this or other Ballot Box Scotland output useful and/or interesting, and you can afford to do so, please consider donating to support my work. I love doing this, but it’s a one-man project and takes a lot of time and effort. All donations, no matter how small, are greatly appreciated and extremely helpful.